The Six Five team discusses Supercomputing 2023.
If you are interested in watching the full episode you can check it out here.
Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.
Transcript:
Daniel Newman: You and I are headed out that way, Pat. I just want to talk a little bit about the evolution of supercomputing before we talk a little bit about themes and stuff we saw. I mean, look, there’s no way in one day that we could get around and hear everything, and there wasn’t there one theme of supercomputing and there wasn’t one announcement of supercomputing, but Pat, very interesting event this year. First and foremost, absolutely jammed wall to wall. You and I got there first night at 7 o’clock PM Mountain time opening, and there was probably 10,000 people lined up at the doors. We were getting shoved over, you and I, the two wafery thin guys we are, we’re getting shoved over by propeller heads, geeks and AI fanatics everywhere that had suddenly returned.
Now again, I talked to some people that had said over the last two or three years … go back two, three years, obviously there was a period of time that that wasn’t going on because shows weren’t going on, but the supercomputing event had gotten thin, it had gotten sparse, it had become very out there and very geeky in the era of AI. And that’s where the FLOPS to TOPS joke came from. We used to measure FLOPS, now we talk about TOPS in case anybody needed an explanation. This was red-hot. And so we saw it wasn’t only the big companies. We had a bunch of Six Five videos with Lenovo. We talked to Lenovo, we talked to Lenovo’s partners, we talked to Imperial College and some of their big users, Pat and we talked to some of their executive team. We talked about everything from liquid cooling to next generation architectures to exponential compute requirements and clusters that had more GPUs than a F1 car has horsepower. Sorry, I had to say that because it’s an F1 week. Yep, you’re wearing it, but I’m living it. I just want to point that out right now.
Patrick Moorhead: I got an invite, but my calendar was too full.
Daniel Newman: Yeah, no, I know something about a sofa, French Bulldogs and I don’t know, fun games.
Anyways, but it was a really, really interesting show. And then there was … so the big companies, the HPEs, the Dells, Lenovos, all very active there. And of course all the silicon, AMDs, NVIDIAs, Intel all on display. And then Pat, there was just this massive ecosystem of what I would call A series all the way to your C and D and E series companies there that were really on big display. And it wasn’t just an HPC show. I want to be very clear, this was an AI show.
There was companies that were in the AI space, GPU powered storage, like our friends at Nyriad that we talked to. You had companies like VAST Data that are building new storage architectures that are powered by AI that were there in a big way on display. You and I both saw the Grok Llama who Grok very focused on accelerated computing language processing units. We were running around spending time. You introduced me to Gopi, from Axiado doing very interesting architectural and security related hardware for building next generation compute network fabrics. So this was just a very interesting show, Pat.
But what I really took away from it, and I know this might be like typical Dan oversimplifying things, but is that the world is really excited about AI and supercomputing is kind of the front edge of it. This is where we’re seeing what all this AI at its maximum deployment with the biggest systems the most cores can do when put to work, to solve problems in healthcare, to solve problems in engineering, to solve problems, even things we talked about with design. This is where that kind of innovation starts, Pat. And so as our friend Pat Gelsinger said something like, the geek is back, this was like the geek is back moment. And by the way, you and I were the coolest guys there because we were bottom third by far of the IQ, but I would put us in the top third on the EQ. But all joking aside, Pat, love to get your takes on what you thought about the event.
Patrick Moorhead: You hit a bunch here, and I do want to give Dan the trademark for FLOPS to TOPS, for those of you who don’t know what a FLOP is, but know what a TOP is. It’s a floating point operations per second. And that was the way that performance was measured in this space for a long time because you were doing visualizations, you were doing simulations, you were doing experiments, you were trying to recreate in the physical world in a digital sense. And then for the past three or four years, AI has plopped in there. And on the machine learning front, HPC experts were using it to narrow down the data set in terms of what the FLOPS needed to work on. And now with generative AI, it’s very different in that they’re actually using generative AI to change the way that they try to solve these problems. And it is truly cool.
Dan, you and I talked with the leaders from Imperial College, the Flatiron Institute with the, I want to call it the Henri system, not the Henry system. We talked to LRZ, the Leibniz Supercomputing Centre leader as well. And they all said-
Daniel Newman: They all had cool names, by the way, I can’t remember.
Patrick Moorhead: They all had cool names and they all said that the AI demands from their users are off the chain and whether that’s their users they serve internally, with the case of Imperial, but also outside like you have with LRZ. Props to HPE, we saw not a ribbon cutting, but a celebration and a toasting. Justin Hotard who runs HPC business unit celebrating the Aurora computer coming in as the number two highest performance, not certified completely yet, but it was only using half of its capability. And if you want to know who’s first, it’s also HPE, the different supercomputer based on AMD technologies. And the one that they were celebrating is based on Intel technology.
It took a long time to get Aurora across the line. In fact, the silicon that was used in there was Knights Landing, which was just completely different architecture that has since been put to bed. And Intel had to create an entire GPU for that system. So it’s good to see it come online, the number one. And number two are still HPE. I want to give the market share leader Lenovo, some kudos in there from a volume standpoint in the top 100. And also thanks to them for letting us interview some of their very important customers and executive.
Yeah, the amount of folks there from national institutes, college institutes, classic supercomputing people, but also like you said, the startups, the Groks, the Nereids, the Axiados of the world. I don’t know if it’s even fair to call VAST to startup given the revenue that they’re cranking out. But their part of the action too is they have a very interesting proprietary way of bundling some of the operating system and the file system actually into the unit itself. And you probably remember that VAST and HPE just did a tie-up, I forget, was it for block storage? One of the capabilities?
Daniel Newman: It was for GreenLake. Yeah.
Patrick Moorhead: Yeah. So interesting stuff.
Author Information
Daniel is the CEO of The Futurum Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise.
From the leading edge of AI to global technology policy, Daniel makes the connections between business, people and tech that are required for companies to benefit most from their technology investments. Daniel is a top 5 globally ranked industry analyst and his ideas are regularly cited or shared in television appearances by CNBC, Bloomberg, Wall Street Journal and hundreds of other sites around the world.
A 7x Best-Selling Author including his most recent book “Human/Machine.” Daniel is also a Forbes and MarketWatch (Dow Jones) contributor.
An MBA and Former Graduate Adjunct Faculty, Daniel is an Austin Texas transplant after 40 years in Chicago. His speaking takes him around the world each year as he shares his vision of the role technology will play in our future.
